THERE ARE TWO BASIC VIEWS OF PROGRAM EVALUATION: (1) A COMPREHENSIVE EFFORT TO DETERMINE THE OBJECTIVES, RESOURCES, PROCESSES, OUTCOMES, AND IMPACT OF A PROGRAM; AND (2) ANY ANALYTICAL ACTIVITY THAT PRODUCES FOR DECISIONMAKING PURPOSES FACTS ABOUT PROGRAM DESIGN AND PERFORMANCE. THE FIRST, MORE RIGOROUS TYPE OF EVALUATION IS APPROPRIATE FOR DEVELOPING OVER THE LONG RUN THE ACCUMULATION OF KNOWLEDGE AND EXPERIENCE THAT SLOWLY MODIFIES THE DECISIONMAKING ENVIRONMENT OF THOSE WHO DETERMINE HOW SOCIAL VALUES ARE TO BE INTERPRETED INTO RESOURCES AND PROCEDURES. TO THE EXTENT THAT THE EVALUATION FUNCTION IS CARRIED OUT AS AN OBJECTIVE PROGRAM RESEARCH AND DOCUMENTATION ACTIVITY, SET APART FROM POLITICAL DECISIONMAKING WITHIN THE MANAGEMENT STRUCTURE, ITS LONG-TERM RELEVANCE IS LIKELY TO BE ENHANCED. THE SECOND TYPE OF EVALUATION INVOLVES THE PROVISION OF INFORMATION TO MANAGERS INVOLVED IN A POLITICAL PROCESS. WHEN GOVERNMENT CONSTRUCTS HIERARCHICAL OFFICES OF EVALUATION, WITH EACH HIGHER OFFICE OVERLOOKING AND USING THE PRODUCTS OF THE LOWER, IT MAKES EVALUATION A THREAT TO MANAGERS AND LIMITS THE EFFECTIVENESS OF EVALUATION AS A TOOL FOR SHORT-TERM DECISIONMAKING. EVALUATION'S SHORT-TERM RELEVANCE IS LIKELY TO BE ENHANCED IF IT IS CONDUCTED AS A LOW-KEY INTELLIGENCE GATHERING EFFORT FOR THE USE OF INDIVIDUAL MANGERS. (LKM)